Note Cards

In the last few weeks the smart money decided that legacy SaaS is dead. Riding the wave of Claude Code, Anthropic hard launched several new tools that took direct shots at software companies serving niche industries and there was lots of selling. The general wisdom is that SaaS profits will be eaten by agents, and that this will happen soon.

I have a different view.

• The software business isn’t about the quality of the software or how well the platform solves problems. These things matter, but like any other company, having skilled operators doing the blocking and tackling of “good business” is what ultimately drives periods of significant momentum and long term growth.
• Successful SaaS companies don’t just minimize customer churn, they turn their customers into champions and evangelists. They build entire communities, host conferences, and work to make their end users part of a tribe. Relationships are a strategic asset, and a team of high-performing CSMs can leverage their customer relationships to drive momentum.
• Switching costs are real. It’s painful to change enterprise systems, and at many enterprises it is a process in the worst possible interpretation of that word. In 2014 I was part of a team tasked to find a new enterprise CRM and EMS to replace a legacy vendor. I left that organization in 2018 with this work well underway, and I recently saw that in 2026 that the new system is finally, fully live.
• Agents have significant, society altering potential. But many of the use cases and solutions aren’t unique. We’re seeing the early signs of commoditization. SaaS companies can leverage their distribution and roll agents and models directly into their platforms. They already have the relationships to glean customer insights, and the workflows in place to turn those learnings into AI-enhanced product features.
• SaaS companies are not going to be caught off guard here. Just listen to any business or tech adjacent podcast and you’ll hear a steady stream of ads for AI-enabled SaaS products. These companies have long seen the future and are making the necessary product moves to retain customers.

This is the moat (and there’s plenty I didn’t mention).

LLMs and agents will have a significant impact on the SaaS business, but I think it’s much more likely that they’ll find themselves as part of an existing ecosystem vs. replacing it outright.

There was a great piece in Bloomberg a couple of weeks back, part news analysis, part tribute, to the NYC MetroCard which was decommissioned at the end of 2025 in favor of a tap and go system that uses NFCs from a passenger’s credit card, digital wallet, or the Metro’s OMNY card.

I am not one to stand in the way of technology. This move makes sense, it’s the same principle as removing TV screens from the seatbacks on planes, or building paths on well-trodden grass in public spaces – organizations should take advantage of how their constituents already live their lives.

What I was struck by in the piece, and had not known about, was the extensive use of MetroCards for promotion of cultural happenings, events, and advertising:

“The MTA has made more than 400 limited edition cards over the past three decades. They include a tribute to celebrate what would have been Brooklyn-born rapper, Notorious B.I.G.’s 50th birthday in 2022 and one displaying the iconic “I Love NY” logo after Hurricane Sandy struck the region and flooded subway stations in 2013.”

These limited edition MetroCards are collector’s items, and can be found on eBay, Etsy, and the New York Transit Museum Store.

I put things like commemorative MetroCards in the category of useful cultural artifacts, the same category as the postage stamp. These have everyday value and utility, but we (via our institutions) have added cultural value on top through design and celebration. Commemorative MetroCards celebrated New York’s wins, its artists, and its culture and through these small artifacts allowed transit riders to take a small and brief moment to celebrate something that can be broadly categorized as a good cultural thing. In this way, people are gently drawn to a bit of a shared identity, or perhaps, just feel a bit of community pride.

Technological advance is a good thing. Here we’re seeing the replacement of the notoriously finicky MetroCards with what should be a more reliable and efficient payment system. We’re also losing something of a cultural piece of NYC. I’d add that as a society, we too often retreat to our phones and move through a world full of other people, and are entirely focused on whatever is on the screen in our hands, and how we’re represented in that world. A change like this, from paper and challenge and shared culture, to the seamless, quick, and solitary, only furthers that juxtaposition.

The NYT recently published a piece about cheap solar panels, and their knock-on effects, rapidly changing the electrical infrastructure, demand, and geo-politics within South Africa. As you might imagine, China is the country building, supplying, and financing the panels, both for residential use, but also for larger utility scale projects.

This is something of a “what have we learned?” moment.

• The United States had early momentum in this space, for a spell a few decades ago we were the leader in solar manufacturing and global supply, but there wasn’t enough of a market.
• In recent years, solar installation was a blossoming and booming industry that was successful in part due to strong federal incentives. These incentives have been lost as part of the “Big Beautiful Bill.” While the US installer industry is still in a good (not great) place, it’s reliant on parts and panels manufactured elsewhere – namely China. This is familiar story, we lost the manufacturing race writ large to China over the course of decades, and solar panels and associated parts are no different.
• The drivers for the solar install demand in the United States are not so different from South Africa. The United States does not have a modern grid, energy prices are rising, and the desire to be “completely independent” is a strong American tradition. To a certain degree, this article feels more like a mirror than we might care to admit.
• The United States has put forth a muddled strategy on energy policy, right at a time when demand is increasing significantly due to AI computing needs. We’re looking for massive increases in power generation (ie. nuclear, which I support) to be the solution, while moving incentives away from decentralized efforts like residential solar and energy storage. That’s not to say that one is better than the other, rather, we need all of these things all at once to create a more stable central grid (the hub), backed by a robust, decentralized model of residential power generation and storage (the spoke).

As a country that is always desiring to exert influence and power globally, we need more rather than fewer arrows in our quiver. What China is able to roll out in South Africa is something that we will not be able to do in the near future, because we can’t even do it in our own backyard. When we get back to the point of understanding that we’ll need residential solar and storage, will we still have a domestic industry that can support the build out?